Rising pressure on data centres from AI is redefining cooling and water management requirements

AI is placing data centres under new levels of strain unlike anything experienced before. At the heart of the challenge, the demands from this scaling technology means traditional cooling approaches are being stretched beyond their practical limits and won’t withstand the stress for much longer. As these energy requirements shift, operators are being forced to rethink how facilities are designed and operated to maintain performance and resilience. 

Liquid cooling technologies are a critical asect of this next phase of growth, offering the efficiency required to support high-density AI environments. However, we’re not talking about a simple upgrade here; the process requires careful planning around infrastructure design, water usage, operational processes and workforce capability. Without a comprehensive and coordinated approach from the start, the drive for capacity risks creating unnecessary delays and operational inefficiencies, all while undermining the steps towards a more sustainable industry already put in place. 

The challenges with water management 

Water, a primary and limited resource, is a major part of the cooling system and is quickly becoming a widely talked-about issue. While proper regulations around this are still developing, investors and customers won’t wait and will ask the big questions: how are these big data centres going to be sustainable if they end up using so much water? From the operator's lens, this means respecting community resources, understanding usage levels and trying to recycle wherever possible. If these basic stakeholder concerns go unheard, data centres will be at a huge reputational and operational risk. 

Thanks to increased power densities, grid constraints and planning delays, these challenges are becoming more complex. UK regulators, for example, are looking at the resilience and sustainability aspects of newer data centres, and similar pressures exist across the whole of Europe. The lesson here is that operators need to think holistically and beyond their immediate capacity needs. In practice, this means that operators should integrate their cooling strategies with water management from the beginning, and facilities should be designed keeping in mind future workloads and resilience, rather than chasing speed to market and risk harming the environment.  

Fulfilling demand for now and the future 

Closed-loop cooling systems are gaining a lot more traction, and for the right reasons. They offer a clear-cut path to sustainability as these systems enable water reuse, reduce overall energy demand, and limit environmental impact. Currently, mid-sized operators are constantly trying to get ahead of their competitors and are facing immense pressure to adopt modern cooling systems at speed while ensuring long-term reliability. The solution here is to partner with experienced delivery teams can help strike the balance between speed, cost, and sustainability.  

Modular construction is also helping operators keep pace with demand. How? Pre-built modules include liquid cooling from the start, which reduces the need for on-site construction time and disruption to local communities. For example, in areas like the Thames Valley or the Midlands, where planning approvals can be slow, modular approaches can allow operators to deploy capacity at speed without compromising sustainability.  

The important part is, modular designs embed sustainability into the build rather than having it as an add-on later. In this day and age, to meet both investor expectations and regulatory changes while still running on schedule, operators need to combine modular deployment with careful water and cooling strategies.  

A new benchmark 

Water and cooling have become vital foundations for the new data centres that will power the world’s growing AI systems. Operators that treat water management and advanced cooling as strategic priorities, not as technical afterthoughts, will strengthen their ability adapt and respond to the increasingly demanding environment around them as it continues to change. This means designing facilities with long-term resilience in mind, embedding efficiency and reuse into cooling infrastructures, and ensuring teams are prepared to manage more complex systems safely and effectively. 

The next generation of AI infrastructure will be judged not just on how quickly capacity can be delivered, but on how responsibly it is sustained. Those that combine collaboration and robust execution will be best positioned to support continued AI innovation while balancing performance and social responsibility.